Feasibility of developing a method of imaging neuronal activity in the human brain: a theoretical review.

Abstract

A theoretical analysis has been performed to suggest directions for research into the development of a device which could image neuronal electrical activity in the human brain in three dimensions. Proposed criteria for the device are a spatial resolution of 1 mm3 and temporal resolution, after averaging to a repeated stimulus, of 1 ms for events related to the action potential, or 1 s for metabolic changes. It is proposed that, for the rapid changes related to the action potential, electron spin resonance using a potential-sensitive spin, label, impedance imaging and NMR are suitable in principle but that only ESR and impedance methods may have sufficient sensitivity and these merit further assessment. For metabolic changes, NMR and PET may be used as at present, and ESR may be developed in time, but images based on these changes would have limited value in that they could only give an indirect, index of neuronal discharge. Unique reconstructions based on the EEG or MEG are theoretically, impossible, and imaging using X-rays, microwaves, or ultrasound may be possible in principle but these techniques would not be sufficiently sensitive.